Package {sportsfeatures}

sportsfeatures package documentation

Description

A synthetic, longitudinal athletic dataset generated through a transparent, rule-based simulation engine. Captures individual activity sessions across multiple athletes, environmental conditions, and physiological responses. Specifically designed as an alternative to legacy teaching datasets by introducing realistic hierarchical repeated measures, complex two-way covariate interactions, and a deliberate Missing Not At Random (MNAR) tracking mechanism suitable for advanced imputation workflows. Methodologies implemented are based on van Buuren (2018) doi:10.1201/9780429492259 and Bates et al. (2015) doi:10.18637/jss.v067.i01.

Author(s)

Maintainer: Mohammad Abbas ma.abbas3107@gmail.com

Authors:

• Mohammad Abbas ma.abbas3107@gmail.com

Access Sports Feature Datasets

Description

A convenient helper function to quickly load and return the package's internal sports features data assets directly into an active variable.

Usage

get_sportsdata(type = c("complete", "missing"))

Arguments

Value

A tibble/data.frame containing the requested sports feature dataset.

Examples

# Get the clean complete dataset
clean_data <- get_sportsdata(type = "complete")

# Get the dataset containing systematic missingness
missing_data <- get_sportsdata(type = "missing")

Comprehensive Sports Features Dataset

Description

Comprehensive Sports Features Dataset

Usage

sports_features

Format

A tibble or data frame with 25 variables describing athlete sessions and performance metrics:

session_id

Unique alphanumeric identifier for each training session.

athlete_id

Unique alphanumeric identifier for each athlete.

datetime

Timestamp of when the training session occurred.

activity_type

Type of exercise performed (e.g., running, cycling, swimming).

region

Geographical area where the session took place.

distance_km

Total distance covered during the session in kilometers.

weather_type

Weather condition during the session (e.g., sunny, rainy, cloudy).

temperature_c

Ambient outdoor temperature in degrees Celsius.

personal_status

Pre-activity physical or mental status reported by the athlete.

is_group_activity

Logical indicator (TRUE/FALSE) if the session was done with a group.

gender

Categorical gender of the athlete.

age

Age of the athlete in years.

base_fitness

Baseline fitness score of the athlete.

base_speed

Baseline average speed capability of the athlete.

base_stamina

Baseline stamina level of the athlete.

base_weight

Baseline body weight of the athlete in kilograms.

resting_heart_rate

Baseline resting heart rate in beats per minute (bpm).

device_type

Type of tracking device used during the session.

speed_kmh

Average speed maintained throughout the session in km/h.

duration_min

Total duration of the training session in minutes.

heart_rate_avg

Average heart rate monitored during the session in bpm.

calories_burned

Estimated total energy expenditure in kilocalories (kcal).

exhaustion_level

Subjective exhaustion level reported after the session.

hydration_status

Hydration level (%) recorded during or after the session.

fatigue_score

Calculated post-activity fatigue accumulation score.

Details

A rich, synthetic sports analytics dataset containing tracking metrics, environmental contexts, physiological markers, and performance data for athletes.

Source

Synthesized sports features analytics framework.

Examples

library(tidyverse)
library(lme4)

# Load the package data
data("sports_features")

# Downsample data for the example to ensure fast execution time (< 2.5s)
demo_data <- head(sports_features, 500)

# ----------------------------------------------------
# DEMO 1: Linear Regression (Fixed Effects)
# Predicting fatigue score based on workload metrics
# ----------------------------------------------------
lm_model <- lm(fatigue_score ~ distance_km + duration_min + speed_kmh + temperature_c,
               data = demo_data)

summary(lm_model)

# ----------------------------------------------------
# DEMO 2: Linear Mixed-Effects Model (Hierarchical MML)
# Controlling for variation across individual athletes (athlete_id)
# ----------------------------------------------------
mml_model <- lmer(fatigue_score ~ distance_km + duration_min + speed_kmh + temperature_c +
                    (1 | athlete_id),
                  data = demo_data)
summary(mml_model)

Comprehensive Sports Features Dataset (With Missing Values)

Description

Comprehensive Sports Features Dataset (With Missing Values)

Usage

sports_features_missing

Format

A tibble or data frame with 25 variables containing structured missing values.

Details

A variant of the core sports analytics dataset containing structured missingness (NA values) across performance tracking columns to demonstrate imputation workflows.

Source

Synthesized sports features analytics framework.